DNA is susceptible to a variety of mutations and chemical modifications. Errors during DNA replication, either mispairing or slippage, result in mismatched base pairs, which occur at a frequency of 10-8 to 10-6. Exposure to UV irradiation or chemical agents may lead to covalently modified DNA bases, and programmed meiotic and mitotic DNA rearrangement, ionizing radiation and oxidative agents can result in double-strand DNA breaks. To maintain genomic integrity and to sustain life, bacteria, archaea and eukarya use conserved mechanisms to repair or to tolerate each type of damage. My research group has continued to carry on structural and functional studies of E. coli and human mismatch repair processes and lesion-bypass DNA synthesis. Mismatch repair (MMR) in E. coli is initiated by three proteins, MutS, MutL and MutH, to specifically target the newly synthesized daughter strand. MutS is an ATPase and recognizes a mismatched base-pair as well as an insertion or deletion of 1-4 nucleotides in one strand. MutH is a latent endonuclease that is both sequence- and methylation-specific; when activated by MutS upon detection of a mismatch, it cleaves 5? to the unmethylated d(GATC) sequence in a hemimethylated duplex. MutL mediates the communication between MutS and MutH, which do not directly interact. Once a nick is introduced to the daughter strand by MutH, UvrD helicase, single-strand binding protein and DNA exonuclease, UvrD are recruited to remove nucleotides from the nick to beyond the mismatch. Homologues of MutS and MutL are found in all eukaryotes, and malfunction of either human MutS or MutL homolog is directly implicated in the susceptibility to hereditary non-polyposis colorectal cancer (HNPCC) and other sporadic cancers. Our previous studies suggest that the broad range of mismatch-repair substrates and high repair specificity are achieved with the high energy factor, ATP, utilized by MutS to verify and proofread mismatch recognition and to recruit MutL to signal for repair. In this year, we have determined the crystal structure of the C-terminal dimerization domain of MutL, characterized its DNA-binding and protein-interacting role in MMR. Based on out biochemical and genetic data, we propose a model that explains how the strand nicking (1st step) occurs either 5' or 3' to the mismatch site and the strand removal (2nd step by UvrD and exonucleases) is usually directed towards the mismatch site. In the past year, we have determined the crystal structures of MutH-DNA complexes with either unmethylated or hemimethylated DNA. We have further characterized how MutL activates MutH and UvrD and how metal ions influences substrate specificity. To understand the relationship between mismatch repair and DNA replication, we have solved the crystal structure of SeqA, a negative regulartor for replication initiation in E. coli and share the same DAN bidning site with MutH. Previously we determined the crystal structure of the DNA-binding domain complexed with its binding site, hemimethyalated GATC. This year, we solve the structure of the polymerization domain of SeqA, which allows us to build a functional SeqA polymer that accounts for its role in DNA replication and segregation. We propose that the competition between the two may regulate the mismatch repair specificity. Lesion-bypass DNA synthesis is carried out by the recently discovered Y-family DNA polymerases, which perform low-fidelity synthesis on undamaged DNA templates and are able to traverse normally replication-blocking lesions, including abasic sites, 8-oxo-G, benzopyrene adducts, and cyclobutane pyrimidine dimers. Y-family polymerases are widespread and enable species from E. coli to human to tolerate UV irradiation and various forms of base modification. Each individual Y-family polymerase exhibits a distinct substrate preference. For example, Pol eta is particularly efficient to bypass the UV crosslinking product, cyclobutane pyrimidine dimers. Mutations in XPV, which encodes human Pol h, are correlated to 20% of xeroderma pigmentosum. After publishing the first Y-family polymerase and DNA complex structure in 2001 and a serier of crystal structures of Dpo4 complexed with a cyclobutane pyrimidine dimers, benzo[a]pyrene adduct, and abasic lesion in 2003 and 2004, this year we cap our studies of Dpo4 by determining its substrate specificity and nucleotide selection. Our structural and biochemical studies suggest that both replicative and translesion DNA polymerase depend on precise metal-ion coordination for the rate-limiting step ? the chemical bond formation. Along the line of metal ion and substate specificity, we have determined the crystal structures of RNase H complexed with an RNA/DNA hybrid substrate. RNase H is an essential enzyme for HIV replication and is the founding member of a large number of endonuclease families. Our crystal structures illustrate how the enzyme recognizes a specific substrate and the metal-dependent hydrolysis mechanism. Based on our structures, we are able to propose a general mechanism for nucleotidyl transfer reactions including DNA transposition and RNAi processing.